Offcourse

Download online courses for offline access – of course! 📚

→ View Documentation & Homepage

Quick Start

# Install
npm install -g offcourse

# Download a course
offcourse sync <course-url>

# Or run without installing
npx offcourse sync <course-url>

Requires Node.js 22+ and ffmpeg for HLS videos. Optional transcription requires the native Cuttledoc CLI on PATH.

Supported Platforms

Platform	URL Pattern
Skool	skool.com/community/classroom
HighLevel	member.*.com/courses/...
Josh Comeau Courses	courses.joshwcomeau.com/<course>
LearningSuite	*.learningsuite.io/student/...
Piccalilli	piccalil.li/<course>/lessons

Key Commands

# Sync a course (auto-detects platform)
offcourse sync <url>

# Sync with options
offcourse sync <url> --skip-videos      # Text only
offcourse sync <url> --dry-run          # Preview
offcourse sync <url> --limit 5          # Test with 5 lessons
offcourse sync <url> --transcribe       # Download, then transcribe with Cuttledoc

# Skool login with community access verification
offcourse login https://www.skool.com/<community>/classroom

# Piccalilli OTP login (sync also prompts automatically)
offcourse login https://piccalil.li/<course>/lessons

# Josh Comeau Magic Link login (sync also prompts automatically)
offcourse login https://courses.joshwcomeau.com/<course>

# LearningSuite login (session is saved per tenant)
offcourse login https://<tenant>.learningsuite.io/student/course/<course>/<id>

# Unlock sequential content (LearningSuite)
offcourse complete <url>

# Configuration (optional)
offcourse config set outputDir ~/Courses  # Default: current directory
offcourse config set videoQuality 720p    # Default: highest
offcourse config set concurrency 3        # Parallel downloads (1-5, default: 2)
offcourse config set extractionConcurrency 6  # Browser tabs (1-8, default: 4)
offcourse config set cuttledocPath /usr/local/bin/cuttledoc
offcourse config set transcriptionLanguage de
offcourse config set transcriptionBackend auto
offcourse config set transcriptionEnhancement local

Transcription

Offcourse integrates Cuttledoc through its native CLI instead of embedding a platform-specific Node module. This keeps course extraction independent from speech engines and lets both tools be installed and released separately.

# Verify the native dependency
cuttledoc --version

# Transcribe new downloads and retry unfinished jobs from earlier runs
offcourse sync <course-url> --transcribe

# Override the configured language, backend, or enhancement for one run
offcourse sync <course-url> --transcribe \
  --transcription-language en \
  --transcription-backend auto \
  --transcription-enhancement off

Transcription is opt-in and sequential by default to keep model memory bounded. For every video, Offcourse writes <video-stem>.transcript.json with Cuttledoc's complete machine-readable result and <video-stem>.transcript.md for reading. SQLite stores the Cuttledoc version, selected settings, durations, attempts, and errors. Repeating --transcribe skips completed videos and retries only unfinished work. --force --transcribe regenerates completed transcripts.

Performance

Course scanning and content extraction use extractionConcurrency browser tabs (default: 4). All tabs share the same authenticated session. Video downloads use a separate concurrency queue (default: 2), so browser work and network/download load can be tuned independently.

Troubleshooting

ffmpeg is missing

Video downloads that use HLS require ffmpeg. Confirm it is available with ffmpeg -version; if the command is missing, install ffmpeg with your operating system's package manager and retry the sync.

Cuttledoc is missing

Transcription is the only feature that requires Cuttledoc. Install a native release from the Cuttledoc repository, verify it with cuttledoc --version, or configure an explicit executable:

offcourse config set cuttledocPath /absolute/path/to/cuttledoc

Playwright cannot find Chromium

Install the browser binary used by the scraper, then retry:

npx playwright install chromium

When working from a repository checkout, use pnpm exec playwright install chromium instead.

A saved session expired

Clear the saved session and force a fresh login for the affected platform URL:

offcourse logout <url>
offcourse login <url> --force

A lesson failed

Offcourse stores per-lesson failure details for every supported platform and supports a targeted retry:

offcourse status <url> --errors
offcourse sync <url> --retry-failed

Development

See CONTRIBUTING.md for commit conventions, Git hooks, and pull-request expectations.

git clone https://github.com/sebastian-software/offcourse.git
cd offcourse
corepack enable
pnpm install
pnpm build
pnpm link --global  # optional: link globally

Commands

pnpm dev               # Watch mode
pnpm check             # Format, lint, types, unit tests, and build
pnpm test              # Unit tests in watch mode
pnpm test:integration  # Network/ffmpeg integration tests

Releases

Releases are automated with Release Please. Merging its release PR updates the package version and changelog, creates the GitHub release, and publishes the package to npm. The recommended setup is npm Trusted Publishing for the publish.yml workflow; an NPM_TOKEN repository secret remains supported as a fallback. A RELEASE_PLEASE_TOKEN secret is recommended so release PRs run the regular pull-request CI. To recover a release after configuring npm authentication, run the Release workflow manually with publish set to true.

Adding a New Platform

1. Create src/scraper/newplatform/ with:
   • auth.ts – Session detection
   • navigator.ts – Course structure
   • extractor.ts – Content extraction
   • schemas.ts – Zod schemas
2. Add a platform sync handler in src/cli/commands/
3. Register URL detection in src/cli/syncPlatform.ts and delegation in src/cli/index.ts

Acknowledgments

Thanks to Sindre Sorhus for the excellent packages powering this project: slugify, conf, delay, execa, ky, p-retry.

License

MIT

---

Open Source at Sebastian Software
Copyright © 2026 Sebastian Software GmbH